Traditional method for data transmission in a communication network is store-and-forward, i.e., nodes other than data sending and receiving nodes do not perform any processing on data contents, and only play the role of a repeater.
The Network Coding theory is a major breakthrough in the field of information theory at the beginning of the 21st century, and its core idea is to perform linear or non-linear processing on received information of each channel in each node of the network, and then forward to downstream nodes, and intermediate nodes or relays play the roles of encoders or signal processors rather than simply amplify-and-forward or decode-and-forward, thus increasing the network throughput. In point-to-multipoint network communication, the node network coding method can be used to make the information transmission rate reach the maximum network traffic, therefore, the network coding is an effective method that can approach the theoretical transmission limit of network capacity and brings a fundamental revolution for the network communication.
Common network coding methods are bit-level XOR network coding, multi-domain network coding, soft information combined network coding and nested network coding, etc.
In the bit-level XOR network coding method, a node performs an XOR operation of the bit level on received data to complete the information combination and achieves the compression of the received data, and this is a most commonly used coding method in which the network coding technology is applied to the wireless communication network.
In the multi-domain network coding method, the multi-domain network coding method constructs user network coded data packets of multiple users, which usually uses the network coding of GF(4), GF(16), GF(64) or GF(256) domain. Compared with the bit-level XOR network coding, the multi-domain network coding can not only support the network coding of multiple users, but also obtain a higher diversity order, which can more effectively improve the performance brought by the diversity gain.
In the soft relay network coding method, the receiving node does not perform a hard decision on the data, but calculates the Log-Likelihood Ratio (LLR) value based on the received soft-bit information after the network coding, and sends the soft-bit information or its variation version to the destination nodes, thus avoiding the hard decision operation and remaining the bit reliability information, and at the receiving end, using the soft-bit signal processing algorithm can better restore the data sent by the source end.
In the nested network coding method, the receiving node receives coded data packets generated by different source nodes, combines the data packets through the nested coding and forwards them to the destination nodes. Because the generated vectors of an encoder are linearly independent with each other, the destination nodes can decode the combined data packets to obtain the data information sent by any source node.
In traditional modes, the source node increases the probability value of a destination node correctly receiving data information by increasing the number of retransmissions. Therefore, the network reliability is usually related to the number of retransmissions. In the wireless communication shown in FIG. 1, when a plurality of source nodes are surrounded by a plurality of destination nodes and the source nodes broadcast data information, how to improve the network reliability is a technical problem to be solved.